The two broadest ways of looking at stars are photometry and spectroscopy. Photometry measures the stars brightness with time, and is a way to detect, for example, eclipsing binaries (two stars whose orbits are aligned so that they regularly cross each other), stars that pulsate, supernova, and so on. These are all photometric measurements. Spectroscopy breaks the starlight up into its colored components (white light is a composite of all the colors) and can measure the chemical elements in the star, how far it has evolved from birth (in most cases), and even how fast it is moving away or toward us (using the Doppler shift in the spectral linessimilar to measuring the shift in pitch of a car horn as the car speeds past you).

The NASA Kepler Mission is going to be able to detect the first Earth-like planets around other stars within this decade. It will do this by measuring very precisely (one part in about 100,000) a drop in the brightness of stars due to a planet orbiting across the disc of the star casting an extremely slight shadow onto the Earth. This "transit" method is the only way presently to detect nearby Earth-sized extrasolar planets, and we have to look at many stars to detect any planets that happen to cross our line-of-sight to produce a transit. But we can only measure this brightness drop for Earth-sized planets if they are crossing the disc of a star the size of the Sun or less. However, half the stars in the region of the Milky Way galaxy where the Kepler spacecraft will be looking (in the constellation Cygnus) are giant or super-giant stars. Thus the problem is how to separate the Sun-like dwarf stars where Earth-like planets are detectable from the giant and super-giant stars where such planets are not detectable?

We could take spectra of the more than 300,000 stars Kepler will be studying, but this could take decades and the launch is set for 2007. We could look at the stars photometricallytaking pictures of whole bunches of them at once (we could get pictures of the whole field of stars in perhaps 6 months) but such usual photometry cannot (generally) tell whether stars are dwarfs or giants. However, it turns out, that we can do a kind of "spectrophotometry" using special filters at the telescope (somewhat like the filters used in color photography but more specialized) to distinguish the dwarfs from the giants while nevertheless taking pictures of the whole region (rather than taking spectra of each star one at a time.)

This approach to cataloguing the Kepler stars (so the spacecraft will know which stars to process and which to ignore) involves selecting special intermediate-band filters that let through a narrow range of select colorsoften centered on spectral lines that give information about the stars size. We can take images of tens of thousands of stars through these filters at a time, thus allowing us to classify hundreds of thousands of stars in just months instead of decades.

Working with colleagues Zoran Ninkov and Robert Slawson at the Rochester Institute of Technology, the method we are presently testing for precision in classifying stars into dwarfs or giants/super-giants is called the Stromvil photometric system, invented by Vutitus Straitzys and colleagues at the Vilnius Observatory in Lithuania. This project would represent the largest and most critical test of this suggested system of photometric stellar classification.

The filters of the Stromvil system are listed as: u, P, v, b, Z, y, and S (yes, the capital and lower-case letters stand for specific filter types). The u-filter is an ultraviolet filtervery hot stars will have a lot more ultraviolet light than red light, for example. The P-filter is on top of the Balmer lines. These are lines of hydrogen that are very sensitive to the size of the star. The v-filter is very sensitive to the scattering of light by interstellar dust. Why is this important? Well if you wanted to know if the Sun was red or yellow (yellow dwarfs are larger than red dwarf stars) but measured it at sunset, you would get a much redder measurement for the Sun than is correct becausein this casethe atmosphere is scattering the blue light away from your eye, and only the red sunset light gets through. This is what happens with interstellar dustit scatters the blue light out into space and our measurements (without the v-filter) would register a too-red star. We might wind up looking at a large blue star thinking it was like the Sun when it is actually "reddened" by the interstellar dust.

The b-filter measures the amount of blue light coming out of the star distinguishing the hot (blue) stars from the cooler (redder) stars. The Z-filter is placed around a spectral line that indicates magnesium and other elements that bunch together at this "color" in cooler Sun-like dwarf stars. It is very sensitive to whether a given cooler star is a dwarf or a giant/super-giant in the same way that the P-filter was sensitive to whether a given hotter star is a dwarf or a giant/super-giant. The y-filter measures the general brightness outside of most spectral lines (called the spectral "continuum") at a color outside of the ultraviolet or bluethat is, in the middle of the yellow region of the spectrum. Thus it can be compared to the b-filter; the cooler stars will appear brighter in the y-filter than in the b-filter, for example. Finally we have the S-filter of the Stromvil color system, which is centered on what is known as the "H-alpha" line, a color region of the stars light output that indicates a number of things, including if the star has a circumstellar shell around ita shell of gas and dust that might indicate that the star is very young, for example (planets may not have been born yet).

Thus over the next few years before launch, we are measuring the color of the stars that the NASA Kepler spacecraft will be looking for planets around. Not only should these observations allow us to tell which stars may best be searched for Earth-like planets, but we should also be able to better imagine what the view from the planets detected might look like too. Perhaps a beautiful orange-double star system in Cygnus is setting on some astronomers getting ready for a nights work and wondering if there is any life around a small yellow star in Orion?